Ci1t MCP
- collapseindex/ci-1t-mcp
CI-1T MCP is an MCP server that exposes 20 tools for prediction stability, ghost detection, drift evaluation, and fleet monitoring.
About
CI-1T MCP connects Collapse Index’s prediction stability engine to Model Context Protocol clients so developers and small ML teams can audit model behavior without a separate dashboard-only workflow. The server advertises twenty tools spanning ghost detection, drift evaluation, and fleet monitoring—aligned with reliability work and late ship gates before wider release. Configure CI1T_API_KEY when you want analyses routed through the hosted API, or use bring-your-own-model paths where the README allows keyless operation. Point CI1T_BASE_URL at your deployment if you self-host. This is an observability and stability integration for agents, not a training framework: register the npm stdio server in Claude Code or Cursor, then invoke tools when production or staging predictions look unstable.
- 20 MCP tools for CI-1T prediction stability workflows
- Detect ghosts, evaluate drift, and monitor model fleets from the agent
- Optional CI1T_API_KEY for backend-routed analysis; BYOM mode supported without key
- npm @collapseindex/ci1t-mcp v1.7.1 with stdio transport
- Configurable CI1T_BASE_URL (default https://ci1t-api.onrender.com)
Ci1t MCP by the numbers
- Data as of Aug 10, 2026 (Skillselion catalog sync)
claude mcp add --env CI1T_API_KEY=YOUR_CI1T_API_KEY --env CI1T_BASE_URL=YOUR_CI1T_BASE_URL ci1t-mcp -- npx -y @collapseindex/ci1t-mcpAdd your badge
Show developers this MCP server is listed on Skillselion. Paste this into your README.
| Package | @collapseindex/ci1t-mcp |
|---|---|
| Transport | STDIO |
| Auth | Required |
| Repository | collapseindex/ci-1t-mcp ↗ |
What it does
Run CI-1T stability checks—ghost detection, drift evaluation, and fleet monitoring—through 20 MCP tools from your coding agent.
Who is it for?
Best when you're running small ML fleets or agent-assisted MLOps and want Collapse Index stability checks without leaving the IDE.
Skip if: Pure prompt-only apps with no measurable prediction pipeline, or teams that do not want external API routing.
What you get
After registration, your agent can call CI-1T-backed stability and drift tools while you debug production ML behavior.
- Twenty CI-1T stability tools available to the coding agent
- Documented API key and base URL configuration for hosted vs BYOM
Recommended MCP Servers
Agent Audit MCPImmutable audit trail for agent-to-agent interactions2
Agentguard47 MCPRead-only MCP server for coding-agent traces, alerts, costs, usage, and budget health.4
Agentic Observability MCPAgent tracing, cost tracking, anomaly detection for LLM agents
Agent Incident Relay MCPAgent Incident Relay MCP — Article 73 5-clock broadcaster.
Agent Infra MCP18 tools: agent memory, API drift detection, and HMAC-signed trust receipts
Agentmonitorrelay MCPAI agent run monitoring with incident replay and SLA receipts.How it compares
ML stability monitoring MCP with 20 tools, not a generic logging agent skill.
FAQ
Who is CI-1T MCP for?
Developers and tiny teams shipping ML-backed features who need drift and ghost checks callable from Claude Code or Cursor.
When should I use CI-1T MCP?
Use it in operate monitoring when predictions drift or fail fleet checks, and during ship perf review before promoting a model.
How do I add CI-1T MCP to my agent?
Install @collapseindex/ci1t-mcp, add the stdio MCP entry, set CI1T_API_KEY if you use hosted analysis and optionally CI1T_BASE_URL, then restart the host.